Induction of long term protective memory will be desirable for most new vaccines including those against pandemic influenza. Our understanding at the cellular and molecular level of factors that determine long term memory is incomplete, and for most viruses natural infection remains the gold standard for inducing life-long protective immunity. Our long term goal is to identify features of inductive immunity that engender long-lived protection, and can be incorporated into vaccine strategies for pandemic and inter-pandemic influenza outbreaks. Our thinking is informed by 1) recent observations on the asymmetric nature of mitosis in initially activated murine T cells, leading to distinct polarized effector and memory progeny;2) current understanding of the importance of concurrent "danger signals" to provide the innate immune system with a context for responding to antigen;3) correlation of effective durable immunity with dual IL-2 and Th1 or Th2 cytokine production within individual cells, and 4) our own observation that one unilateral injection of killed influenza vaccine was able to strongly activate contralateral lymph nodes in a healthy normal patient with a history of natural infection several years prior to vaccination. Taken together, these observations raise the possibility that antigen specific replication in contralateral nodes can be seen following unilateral vaccination with non- replicating vaccine, due to presentation by DCs of antigen retained from past infections (or vaccinations). Further, it is possible that the quality of cellular response to such "archived" antigen, presented at relatively low concentrations, and in the absence of concurrent danger signals, may differ from the response to antigen from adjuvanted vaccine or active infection. Our immediate objective is to identify key aspects of natural infection with influenza vs. killed trivalent vaccine (FluShield) vs. attenuated live vaccine (FluMist), that influence the quality of priming for subsequent restimulation by unilateral FluShield, one or more years later. In particular, we will test the specific hypotheses that natural infection and live attenuated virus vaccination, but not repeated injection of FluShield into the same arm, will prime the host for a specific activation of spleen and bilateral lymph nodes following a subsequent i.m injection of FluShield, as detected by FDG-PET. We are intrigued by the possibility that T cells proliferating in contralateral nodes containing low concentrations of Ag in the absence of vaccine adjuvant delivered danger signals will develop a different phenotype and cytokine profile than T cells proliferating in the presence of vaccine derived danger signals. Results could lead to improved strategies for the administration and design of new vaccines, capable of inducing better long term immunity.